Build Layer Image Normalization for Additive Manufacturing Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Additive manufacturing apparatuses face challenges due to sensor variations across the build plate, leading to data processing issues and potential errors during the build process, such as inconsistencies, gas pockets, or improper fusing, which can be exacerbated by complex data and large file sizes.
Innovation Solution
The method involves obtaining images of build layers, removing variations between data points, normalizing features to eliminate location dependence, and adjusting energy beam parameters based on normalized features using a machine learning algorithm to generate reduced datasets, thereby mitigating errors and improving data processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple sensors are used to monitor the build process, then measurement precision is improved, but device complexity increases and data processing becomes more difficult
Solution Approach 1:
The patent combines multiple sensor data streams into a unified image-based monitoring system. Instead of processing separate data from multiple sensors, the system captures build layer images that integrate information about powder distribution, melt pool characteristics, and build quality in a single coherent dataset, reducing system complexity while maintaining monitoring precision.
Solution Approach 2:
The system creates visual copies (images) of the build layer that represent the physical state of the build plate. These images serve as simplified representations that capture essential build process information without requiring complex processing of raw sensor data from multiple sources, making data analysis more manageable.
2Measurement precision
If multiple sensors generate detailed data, then measurement precision is improved, but productivity decreases due to slow data processing
Solution Approach 1:
The system extracts only the most critical features from build layer images, such as melt pool geometry, powder distribution patterns, and defect characteristics. By focusing on these key features rather than processing all raw pixel data, the system maintains high measurement precision while significantly reducing data processing time and preserving build productivity.
Solution Approach 2:
The patent segments the build layer image into distinct regions of interest (e.g., melt pool area, powder bed, already solidified layers) and processes each segment independently. This segmentation allows parallel processing of different build zones, reducing overall data processing time while maintaining comprehensive monitoring coverage.
3Measurement precision
If sensors monitor the entire build plate, then measurement precision is improved, but data complexity increases leading to processing delays
Solution Approach 1:
The system applies different analysis methods to different regions of the build plate based on local characteristics. For example, the melt pool region receives detailed geometric analysis while already solidified areas receive simpler defect detection. This localized approach maintains comprehensive spatial monitoring while reducing overall data processing complexity through region-specific optimization.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
An additive manufacturing apparatus (200), a computing system (1210), and a method for operating an additive manufacturing apparatus (200) are provided. The method includes (1010) obtaining two or more images corresponding to respective build layers at a build platform (210), wherein each image comprises a plurality of data points comprising a feature and corresponding location at the build platform (210); (1020) removing variation between the features of the plurality of data points; and (1030) normalizing each feature to remove location dependence in the plurality of data points.